There are lies, damned lies, and statistics. Statistics can be used to prove anything. For instance, if last year one person died from a fatal computer keyboard explosion and this year two people suffered that fate, you could truthfully trumpet the news that "Fatal Keyboard Explosions Increas 100 Percent!"
So, when creationists scream that the odds are a trillion-to-one that two inert chemicals would form the basics of life, have they taken into account that literall tens of trillions of chemical interactions are going on in any given year? No. They let the trillion-to-one claim stand without actually checking the math. Now, for someone like gore3000, who is nothing more than a simple computer program, checking the math never crosses his mind. However, most creationists aren't nearly as one dimensional as the aforementioned algorythm, so one would expect there to be at least a little questioning of any particular paradigm (this is a tenet of science, so it may be hard for some dyed-in-the-wool creationists to accomplish).
Some ardent creationists have begun questioning their arguments against evolution. The most notable example is Answers in Genesis: Arguments We Think Creationists Should Not Use. While it does not address the statistics issue, it does puncture a number of the hoary canards used by creationists on these threads.